Sharwood, RE, Ghannoum, O, Kapralov, MV, Gunn, LH and Whitney, SM

Temperature responses of Rubisco from Paniceae grasses provide opportunities for improving C3 photosynthesis.

http://researchonline.ljmu.ac.uk/id/eprint/5501/

Article

LJMU has developed LJMU Research Online for users to access the research output of the University more effectively. Copyright © and Moral Rights for the papers on this site are retained by the individual authors and/or other copyright owners. Users may download and/or print one copy of any article(s) in LJMU Research Online to facilitate their private study or for non-commercial research. You may not engage in further distribution of the material or use it for any profit-making activities or any commercial gain.

The version presented here may differ from the published version or from the version of the record. Please see the repository URL above for details on accessing the published version and note that access may require a subscription.

For more information please contact researchonline@ljmu.ac.uk

http://researchonline.ljmu.ac.uk/

Citation (please note it is advisable to refer to the publisher’s version if you intend to cite from this work)

Sharwood, RE, Ghannoum, O, Kapralov, MV, Gunn, LH and Whitney, SM (2016) Temperature responses of Rubisco from Paniceae grasses provide opportunities for improving C3 photosynthesis. Nature Plants, 2. ISSN 2055-0278

LJMU Research Online

Variation in response of C3 and C4 Paniceae Rubisco to temperature 1

provides opportunities for improving C3 photosynthesis 2

3

Robert E. Sharwood1+, Oula Ghannoum2+*, Maxim V. Kapralov3, Laura H. Gunn1,4, and 4

Spencer M. Whitney1+*. 5

6

1Research School of Biology, Australian National University, Canberra ACT, 2601, 7

Australia. 8

2 Hawkesbury Institute for the Environment, Western Sydney University, Richmond, 9

New South Wales, 2753, Australia. 10

+ ARC Centre of Excellence for Translational Photosynthesis, Australian National 11

University Canberra ACT, 2601, Australia. 12

3Current Address: School of Natural Sciences and Psychology, Liverpool John Moores 13

University, Liverpool, L3 3AF, United Kingdom 14

4Current address: Department of Cell and Molecular Biology, Uppsala University, 15

Uppsala, SE-751 24, Sweden. 16

*Corresponding Authors: o.ghannoum@westernsydney.edu.au and 17

spencer.whitney@anu.eud.au 18

Abstract: 19

Enhancing the catalytic properties of the CO2-fixing enzyme Rubisco is a target for 20

improving agricultural crop productivity. Here we reveal high diversity in the kinetic 21

response between 10°C to 37°C by Rubisco from C3- and C4-species within the grass tribe 22

Paniceae. The CO2-fixation rate (kcatC) for Rubisco from the C4-grasses with NADP-malic 23

enzyme (NADP-ME) and phosphoenolpyruvate carboxykinase (PCK) photosynthetic 24

pathways was two-fold greater than the kcatC of Rubisco from NAD-ME species over all 25

temperatures. The decline in the response of CO2/O2 specificity with increasing 26

temperature was slower for PCK and NADP-ME Rubisco – a trait which would be 27

advantageous in the warmer climates they inhabit relative to the NAD-ME grasses. 28

Variation in the temperatures kinetics of Paniceae C3-Rubisco and PCK-Rubisco were 29

modelled to differentially stimulate C3-photosynthesis above and below 25°C under current 30

and elevated CO2. Identified are large subunit amino acid substitutions that could account 31

for the catalytic variation among Paniceae Rubisco. Incompatibilities with Paniceae 32

Rubisco biogenesis in tobacco however hindered their mutagenic testing by chloroplast 33

transformation. Circumventing these bioengineering limitations is critical to tailoring the 34

properties of crop Rubisco to suit future climates. 35

Concerns about how escalating climate change will influence ecosystems are particularly 36

focused on the consequences to global agricultural productivity where increases are 37

paramount to meet the rising food and biofuel demands. Strategies to improve crop yield 38

by increasing photosynthesis have largely focused on overcoming the functional 39

inadequacies of the CO2-fixing enzyme Rubisco. A competing O2-fixing reaction by 40

Rubisco produces a toxic product whose recycling by photorespiration consumes energy 41

and releases carbon. The frequency of the oxygenation reaction increases with temperature. 42

To evade photorespiration many plants from hot, arid ecosystems have evolved C4 43

photosynthesis that concentrates CO2 around Rubisco that also facilitates improved plant 44

water, light and nitrogen use. Here we show extensive catalytic variation in Rubisco from 45

Paniceae grasses that align with the biochemistry and environmental origins of the different 46

C4 plant subtypes. We reveal opportunities for enhancing crop photosynthesis under 47

current and future CO2 levels at varied temperatures. 48

The realization of the dire need to address global food security has heightened the need for 49

new solutions to increase crop yields1. Field tests and modelling analyses have highlighted 50

how photosynthetic carbon assimilation underpins the maximal yield potential of crops2. 51

This has increased efforts to identify solutions to enhance photosynthetic efficiency and 52

hence plant productivity3. Particular attention is being paid to improving the rate at which 53

ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco, EC 4.1.1.39) can fix CO2 54

(refs 4–8). The complex structure and catalytic chemistry of Rubisco has so far made 55

improving its performance difficult9–11. Diversity screens have identified natural Rubisco 56

variants with catalytic improvements of potential benefit11–17, but most overlook the 57

influence of broad changes in temperature. 58

In C3-plants Rubisco performance is hampered by slow CO2-fixation rates (kcatC 59

~2-3 s-1) and competitive O2-fixation that produces 2-phosphoglycolate, which requires 60

recycling by the energy-consuming and CO2-releasing photorespiratory cycle. This 61

necessitates C3-plants invest up to 50% of their leaf protein (~25% of their nitrogen) into 62

Rubisco to sustain viable CO2 assimilation rates1-3. A reduction in the atmospheric CO2:O2 63

ratio during the Oligocene period (~30 million years ago) heightened plant photorespiration 64

rates, particularly in hot, arid environments4. This led to the convergent evolution of C4 65

photosynthesis along >65 multiple independent plant lineages5. C4-plants contain a CO2 66

concentrating mechanism (CCM) that allows Rubisco in the chloroplasts of bundle sheath 67

cells (BSC) to operate under near-saturating CO2 levels. This supresses O2-fixation and 68

photorespiration. The BSC CCM begins in the adjoining mesophyll cells (MC) where 69

inorganic carbon, as HCO3-, is fixed to phosphoenolpyruvate (PEP) by PEP carboxylase 70

(PEPC) to form the C4-acid oxaloacetate (OAA). Conversion of OAA to malate (or 71

aspartate) precedes its diffusion into the BSC where it is decarboxylated to elevate CO2 72

around Rubisco. The three biochemical subtypes of C4-plants correlate to the dominant 73

decarboxylation enzyme: nicotinamide adenine dinucleotide (NAD) phosphate malic 74

enzyme (NADP-ME), NAD malic enzyme (NAD-ME) or phosphoenolpyruvate 75

carboxykinase (PEP-CK)6,7. 76

An escalating appreciation of the significant kinetic variation among plant, algae 77

and prokaryotic Form I Rubisco has, until recently, paid little consideration to the 78

functional diversity and potential of C4-Rubisco to improve C3-photosynthesis 8. 79

Adaptation of C4-Rubisco to elevated BSC CO2 has beneficially increased carboxylation 80

rate (kcatC) but unfavourably lowered CO2-affinity (i.e. increased Km for CO2). The increase 81

in kcatC endows C4-plants with accompanying improvements in their nitrogen (less Rubisco 82

required), water (reduced stomata apertures needed) and energy (reduced photorespiration) 83

use efficiencies9 – features considered of potential benefit to engineering in C3-plants10. 84

What remains unclear is the extent to which variation in the ancestral timing, CCM 85

biochemistry and biogeographical origin has influenced the kinetic evolution C4-Rubisco, 86

its response to temperature and it’s potential to benefit C3-photosynthesis without a CCM. 87

Here we examine the diversity in the temperature response (10°C to 37°C) of 88

Rubisco catalysis in Paniceae grasses comprising species with C3, C3-C4 intermediate (C2) 89

and all three C4 biochemical subtypes. We identify significant variation in the kinetic 90

properties of Paniceae Rubisco which correlates with the photosynthetic physiology and 91

environmental distribution of each species. We show by modelling how the potential of 92

Paniceae Rubisco to differentially improve C3-photosynthesis at low and high temperatures 93

under current and future CO2. Differences in the chaperone requirements of monocot 94

Rubiscos are revealed that prevent use of chloroplast transformation to validate Paniceae 95

Rubisco “catalytic switches” using the surrogate model dicot plant tobacco. 96

Materials and Methods 97

Plant Seeds and Growth Conditions 98

Seeds for Panicum antidotale, P. monticola, P. virgatum, P. milliaceum, P. coloratum. P. 99

deustum, P. milioides, P. bisulcatum, Megathyrsus maximus, Urochloa panicoides, U. 100

mosambicensis, Cenchrus ciliaris, Setaria viridis and Steinchisma laxa were obtained 101

from Australian Plant Genetic Resources Information System (QLD, Australia) and 102

Queensland Agricultural Seeds Pty. Ltd., (Toowoomba, Australia) (Table S1) and sown in 103

germination trays containing a common germination mix. Three to four weeks after 104

germination, three seedlings were transplanted into 5L pots containing potting mix and 105

grown in the glass house under natural illumination at 28°C/22°C D/N. Plants were watered 106

regularly with the addition of a commercial of liquid fertilizer (General Purpose, Thrive 107

Professional, Yates, Australia). 108

Leaf dry matter carbon isotope composition 109

Leaf dry matter carbon isotope composition was determined to confirm which species use 110

the C4 photosynthetic pathway. Leaf discs were oven-dried then combusted in a Carlo Erba 111

elemental analyser (Model 1108, Milan, Italy). The emitted CO2 was analyzed by mass 112

spectrometry (VG Isotech, Manchester, UK) and the δ13C was calculated as [(Rsample–113

Rstandard)/ Rstandard]*1000, where Rsample and Rstandard are the 13C/12C ratio of the sample and 114

the standard Pee Dee Belemnite (PDB), respectively 115

Rubisco catalytic measurements 116

Rates of 14CO2 fixation by fully activated Rubisco were measure at 10 to 37°C using 117

soluble leaf protein extracted from 0.5 to 2.0 cm2 of leaf material extracted in 1 mL 118

extraction buffer as described by Sharwood et al (2008)11. Preliminary assays as described 119

in Sharwood et al., (2016) 12 were used to confirm the suitability of the extraction process 120

for sustained maximal activity over 30 min at 25°C. The 14CO2 fixation assays (0.5 mL) 121

were performed in 7-mL septum-capped scintillation vials in reaction buffer (50 mM 122

EPPES-NaOH (pH 8.19 at 25°C), 10 mM MgCl2, 0.4 mM RuBP) containing varying 123

concentrations of NaH14CO3 (0–40 μM) and O2 (0–30%) (vol/vol), accurately mixed with 124

nitrogen using Wostoff gas-mixing pumps. The vials were incubated at the appropriate 125

assay temperature for at least 1 hr before adding 20 µL of the soluble leaf protein to initiate 126

the reaction. The assays were terminated with 0.1 mL of 20% (v/v) formic acid after 1 min 127

(for the assays at 25 to 37°C) or 2 min (for the assays at 10 to 20°C). The Rubisco kinetic 128

measurements were performed using two to six biological samples (see Table S2 for 129

detail). Each protein sample was assayed in duplicate following incubation at 25°C for 8 130

and 12 min. The carboxylation activity varied by <2% between each technical replicate. 131

This confirmed each Rubisco was fully activated after incubating 8 min at 25°C with no 132

detectable loss of activity after incubating a further 4 min. The Rubisco content 133

(determined by 14C-CABP binding12) and integrity of the extracted L8S8 holoenzyme 134

Rubisco was confirmed by non-denaturing PAGE 13. For each experiment a soluble leaf 135

protein preparation was added to four assays containing the highest [14CO2] and 5 nmol of 136

purified RuBP. After reacting to completion (1 to 3 h at different temperatures) they were 137

treated with formic acid, dried and processing for scintillation counting. The measured 14C 138

cpm in each assay varied by <0.5% and the average value divided by 5 to derive the 14CO2 139

specific activity. The values for pH, pK1, pK2 and q (the CO2 solubility at 1 atm) used to 140

calculate CO2 levels in the assays at the different temperatures are provide in Table S5. 141

The Michaelis-Menton constants (Km) for O2 (KO), for CO2 under nitrogen (KC) or air levels 142

of O2 (KC21%O2) were determined from the fitted data. The maximal rate of carboxylation 143

(Vcmax) was extrapolated from the fitted data and the caboxylation rate (kcat

c) derived by 144

dividing Vcmax by the Rubisco-catalytic site content quantified by [14C]-2-CABP binding 145

11. 146

Rubisco CO2/O2 specificity (SC/O) was measured using Rubisco rapidly purified by ion 147

exchange then Superdex 200 (GE Life Sciences) size exclusion column chromatography 148

13. The assays were equilibrated with 500 ppm CO2 mixed with O2 using Wostoff gas-149

mixing pumps and SC/O calculated using CO2:O2 solubility ratios of 0.033, 0.035, 0.036, 150

0.038, 0.039 and 0.041 at assay temperatures of 10, 15, 20, 25, 30 and 35°C, respectively 151

(see Table S5 for gas solubility detail). 152

rbcL amplification, sequencing and phylogenetic alignment 153

Replica genomic DNA preparations (2 to 4) from each grass species were purified from 154

~0.5 cm2 leaf discs using a DNeasy Plant Mini Kit (Qiagen) according to manufacturer’s 155

instructions. The full length rbcL coding sequence (including adjoining 5′UTR and 3′UTR 156

sequence) was PCR amplified from each DNA preparation using primers 5′PanrbcL (5′- 157

CTAATCCATATCGAGTAGAC -3′) and 3′PanrbcLDNA (5′- 158

AGAATTACTGCATTTCGTAAC -3′). The amplified products varied in size between 159

species (1504 to 1589–bp) but each showed identical sequence for the independent DNA 160

preparations from each species. DNA sequences were translated into protein sequences and 161

aligned using MUSCLE (Edgar, 2004) and the rbcL phylogeny reconstructed using 162

maximum-likelihood inference conducted with RAxML version 7.2.6. 163

Chloroplast transformation of Panicum rbcL into tobacco 164

Plasmids pLevPdL and pLevPbL were biolistically transformed into the plastome of the 165

tobacco genotype cmtrL1 as described 13 to derive the transplastomic genotypes tobPdL and 166

tobPbL that, respectively, coded the rbcL genes from P. deustum and P. bisulcatum (in 167

addition to the aadA gene coding spectinomycin resistance) in place of the tobacco rbcL 168

gene. RNA blot, [14C]-2-CABP binding and PAGE analyses of Rubisco expression were 169

performed on independent homoplasmic lines for each genotype as described above with 170

additional experimental detail provided in Figure S7. 171

Statistical analysis 172

Statistical analysis was carried out using one-way (species or photosynthetic type/ subtype) 173

or two-way analysis of variance, ANOVA (Statistica, StatSoft Inc. OK, USA). Means were 174

grouped using a Post-hoc Tukey test. Detailed description of the temperature response 175

analysis and modelling are provided in the Figure and Table legends for convenience. 176

Results 177

Our comprehensive evaluation of Rubisco kinetics within the Paniceae tribe included two 178

C3, one C3-C4 intermediate (signified C27), four NADP-ME, four PCK and three NAD-ME 179

species (Table S1). Rubisco from tobacco, our model plant for Rubisco engineering and 180

that commonly used in biochemical modeling, was included as a control. The C3 and C4 181

physiologies of each species were confirmed using dry matter carbon isotope ratio (δ13C) 182

measurements (Table S1). As expected, the δ13C kinetic isotope effect was significantly 183

lower in the C2 and C3 species (≈ -28.7‰) relative to the C4 specie (≈ -13.3 to -14.6‰) 184

(Fig 1a). 185

The carboxylation properties of Paniceae Rubisco synchronize with C4-subtype. 186

Substantial variation was found in the Rubisco kinetics measured at 25°C among enzymes 187

from C2/C3-species and each C4-subtype (Table S1). Relative to the carboxylation rates 188

(kcatc) of the C2/C3 species, the Rubisco kcat

c was marginally higher in NAD-ME and 2-fold 189

greater in the NADP-ME and PCK species (Fig. 1b). Consistent with the co-dependency 190

of KC and kcatc 14, greater reductions in CO2-affinity (i.e. higher KC’s) were found for 191

Rubisco from NADP-ME and PCK species relative to the NAD-ME and C2/C3 species (Fig 192

1c). Less variation was observed for the averaged oxygenation rates (kcatO; Fig. 1d) and O2-193

affinities (KO; Fig. 1d) among C3 and C4 Rubisco. Nevertheless, the NADP-ME Rubiscos 194

tended to show less sensitivity to O2 inhibition (i.e. a higher KO). This improvement did 195

not, however, improve the specificity for CO2 over O2 (SC/O) of NADP-ME Rubisco, which 196

was significantly lower than the more similar SC/O of Rubisco from the C3, NAD-ME and 197

PCK species (Fig 1f). 198

Analysis of these core catalytic parameters underscored how the strong positive 199

correlation between kcatc and KC shared by plant Rubisco 14-21 extends to Paniceae C3- and 200

C4-Rubisco (Fig 1g). Uniquely, each C4-subtype Rubisco aggregated at a distinctive 201

position along the regression indicative of adaptation to the differences in CCM 202

efficiencies and biogeography among C4-subtypes22, or reflective of differences in resource 203

partitioning to Rubisco that, in NADP-ME plants for example, correlate with improved N-204

use efficiency9. The “subtype-grouping” of the carboxylase kinetics was not evident in the 205

increasingly weaker linear correlations between kcato and KO (Fig 1h), kcat

c and kcato (Fig 1i) 206

and KC with KO (Fig 1j). Evidently, the coordinated changes in kcatc and KC for each 207

Paniceae C4-subtype are not tightly coupled to changes in oxygenase kinetics. This feature 208

is common to Rubisco due to differences in the mechanism and energy profiles of the 209

carboxylation and oxygenation reactions, a property that has facilitated the evolution of 210

diverse Rubisco kinetics 14,23. 211

The potential for Paniceae Rubisco to improve C3-photosynthesis at 25°C. 212

A recent study of Rubisco kinetic diversity revealed how the enzyme from some C4-213

species, such as the increased SC/O and carboxylation efficiency under ambient O2 (kcatc / 214

KC21%O2) of Zea mays (maize) NADP-ME Rubisco, has the potential to improve C3-215

photosynthesis8. The bi-functionality of Rubisco necessitates consideration of both O2 and 216

CO2-fixing activities when evaluating improvement within C3-photosynthesis1,24,25, and 217

does not necessarily accord with a higher kcatc1,8. A correlative analysis of these parameters 218

for Paniceae Rubisco identified a weak relationship between kcatc and SC/O (r2 = 0.43; Fig 219

2a) supporting mounting evidence that the trade-off proposed between these parameters14,21 220

shows significant natural divergence18,26. Differences in O2 inhibition among the Paniceae 221

Rubisco (i.e. variable KO values, Fig 1e) resulted in KC21%O2 values (quantified as 222

KC(1+[O2]/KO)) that showed a weaker co-dependence with kcatc (r2 = 0.76; Fig 2b) relative 223

to KC (r2 = 0.88; Fig 1g). This underscores the inaccuracy of using KC measures as a proxy 224

to interpret the relative CO2-affinity of Rubisco under ambient O2 (i.e. KC21%O2). 225

The biochemical models of Farquhar et. al., (1980)24 provide a useful tool to 226

evaluate how the kinetic properties of Rubisco influence carbon assimilation in C3-plants. 227

These C3-models often use tobacco Rubisco as the reference 1,8,27,28. This stems from 228

tobacco having well characterized Rubisco kinetics, it being the model species for 229

bioengineering Rubisco by chloroplast and nucleus transformation, and its potential to 230

support higher rates of photosynthesis at 25°C under low chloroplast CO2 pressures (Cc) 231

than wheat Rubisco8,29. Figure 2c shows comparable C3-modeling using the averaged SC/O, 232

kcatc / KC

21%O2 and kcatc values from each Paniceae biochemical subtype (Table S1). Under 233

low Cc where CO2-assimilation rates are carboxylase limited, Rubisco from the NADP-234

ME and PCK species would support higher rates of photosynthesis than the Paniceae C3 235

and NAD-ME and tobacco Rubisco (Fig 2c). Under higher Cc where photosynthesis 236

becomes limited by light dependent rates of electron transport the lower SC/O of the 237

Paniceae C4-Rubiscos would support lower rates of CO2-assimilation relative to tobacco. 238

In contrast the higher SC/O of Paniceae C3-Rubisco would enhance their CO2-assimilating 239

capacity at Cc’s above ~240 µbar (Fig 2c). 240

The temperature diversity of Paniceae Rubisco 241

Most diversity screens of Rubisco kinetics are undertaken at 25°C and possibly one or two 242

other temperatures18,20,30,31. More rigorous studies providing kinetics that can be 243

extrapolated over a broad temperature range have primarily focused on Rubisco from C3-244

plants 15,19,28,32,33. In general, the level of kinetic variation has been sufficient to highlight 245

weakness in the customary use of the temperature response for tobacco Rubisco kinetics 27 246

to reliably model the photosynthetic responses of other species. This weakness is 247

particularly apparent from our high precision temperature response measurements that 248

reveal substantial kinetic diversity among Paniceae Rubisco from NAD-ME, NADP-ME, 249

PCK and C2/C3 groupings (Fig 3). The parameters analyzed were SC/O, kcatc and KC

21%O2 250

(averting the need to measure KO for C3-modeling purposes) at six incremental 251

temperatures between 10 and 37°C (Fig S1 to S3). The activation energies (ΔHa) for each 252

Rubisco parameter were comparable among the Paniceae species tested within each C3 and 253

C4-subtype grouping (Table S3). This facilitated the derivation of averaged ΔHa and scaling 254

constant values (c) for each parameter (Fig 3a). Consistent with the highly variable 255

properties of Rubisco from each Paniceae grouping (Fig 1) the ΔHa values showed greater 256

variation (Fig 3a) than that reported for Rubisco from differing C3 species18 and C4-dicot 257

Flaveria species19. This divergence is readily apparent from plots using the averaged ΔHa 258

values to extrapolate the temperature response of kcatc (Fig 3b), KC

21%O2 (Fig 3c), kcatc / 259

KC21%O2 (Fig 3d) and SC/O (Fig 3e) for each Paniceae Rubisco grouping and tobacco 260

Rubisco (control). 261

The kcatc for each Rubisco showed a biphasic Arrhenius temperature response above 262

and below ~25°C (Fig S1). This necessitated the derivation of two ΔHa and c 263

measurements for each Rubisco kcatc (Fig. 3a) whose modeled temperature responses 264

intersect at 25°C (Fig 3b). Importantly, the dual activation energy response of kcatc is 265

universal to all temperature response studies of plant Rubisco but mostly not 266

acknowledged15,18-20,28,30-32. The basis for the asymmetric response remains uncertain. 267

At each assay temperature the kcatc and KC

21%O2 for each NADP-ME and PCK 268

Rubisco were consistently ~2-fold higher than Rubisco from P. bisulcatum (C3), P. 269

milioides (C2) and each NAD-ME species (Table S1 and S2). The shared change in kcatc 270

with temperature by NAD-ME and C2-C3 Rubisco (Fig 3b) was not evident in the measured 271

KC21%O2 values that showed a heightened rate of increase with temperature by NAD ME 272

Rubisco (Fig. 3c). The biphasic response of kcatc was evident in corresponding measures of 273

carboxylation efficiency (kcatc / KC

21%O2) that showed two linear responses that deviated at 274

temperatures above and below ~25°C for each Rubisco (Fig 3d). A comparable kcatc / KC

275

temperature dependency is apparent for the Rubisco from Flaveria C3 and C4 species19 and 276

Setaria viridis C4-Rubisco15. The differential slopes of the linear regression underscores 277

the significant variation in kcatc and KC

21%O2 between each Paniceae Rubisco grouping (both 278

below and above 25°C) and emphasizes the extrapolative limitations of kinetic surveys 279

examining only a few temperatures. This is particularly relevant for measures of SC/O where 280

the extent of exponential change appears more prevalent with reducing temperature (Fig. 281

3e). 282

The potential for improving C3-photosynthesis under current and future CO2 and 283

elevated temperatures 284

The temperature response of each Paniceae Rubisco showed varying extents of 285

improvement in SC/O and/or kcatc/KC

21%O2 relative to tobacco Rubisco (Fig S3 and S4). The 286

improvements observed were greater for Rubisco from P. bisulcatum (C3), Urochloa 287

panicoides (C4-PCK) and P. deustum (C4-PCK). When modeled in a C3-photosynthesis 288

context under varying temperature and chloroplast CO2 pressures (Cc) under saturating 289

illumination (Fig S5) all three Rubiscos differentially improved carbon assimilation 290

relative to tobacco Rubisco (Fig 4). At temperatures below 20°C the simulated 291

photosynthesis rates were limited by electron transport rate at atmospheric CO2 (Ca) levels 292

above those of pre-industrial times (Ca >280 ppm ≈ Cc >170 ppm) (Fig S5). Improvements 293

in SC/O were therefore required to enhance photosynthetic rates at low temperature, a 294

kinetic trait afforded by Paniceae C3/C2 Rubisco (Fig 3c), in particular P. bisulcatum 295

Rubisco (Fig 4 and S3). However, the heightened SC/O sensitivity of P. bisulcatum Rubisco 296

to increasing temperature (Fig S3a) caused these improved photosynthetic rates to wane 297

with increasing Ca and temperature (Figs 4 and S5). In contrast, the improved SC/O response 298

to temperature by P. deustum Rubisco (Fig S3) and rising kcatc/KC

21%O2 (Fig S4) 299

substantially improved photosynthesis rates at temperatures >20°C under current and 300

future Ca levels (Fig 4b). This improvement exceeded that simulated for U. panicoides 301

Rubisco whose lower SC/O hindered its enhancement potential. The antagonistic advantage 302

of these Paniceae Rubisco to lower (P. bisulcatum) and higher (U. panicoides, P. deustum) 303

temperatures were not apparent from the 25°C kinetic measurements. 304

The challenge of identifying catalytic switches in Paniceae Rubisco 305

The rbcL gene in the plastome of each Paniceae species were fully sequenced and their 306

amino acid sequences compared (Fig 5a). A phylogenetic analysis revealed the L-subunit 307

sequences branched according to C3 and C4-subtype physiology (Fig. S6) except P. 308

monticola (NADP-ME) and M. maximus (PCK) Rubisco that share identical L-subunits 309

but show large catalytic variation (Table S1 and S2). This suggests that Paniceae Rubisco 310

small subunits influence catalysis, a function likely shared by the small (S-) subunits of 311

sorghum34 and wheat35 Rubisco. 312

While examination of the S-subunit diversity among Paniceae remains to be 313

undertaken, our L-subunit analysis identified Ala 94 and Ala 228 (spinach Rubisco 314

numbering) as exclusive to C4 Rubisco with Ser 328 and Glu 470 substitutions favored by 315

PCK and NADP-ME Rubisco (Fig 5a). Potential roles for amino acids 94 and 228 in 316

catalysis are unclear. Residue 94 is distal to the catalytic sites in the equatorial region of 317

Rubisco exposed to solvent where it facilitates interactions with Rubisco activase 318

(RCA)36,37. Residue 228 is within the α2 helix also distal to the catalytic site but proximal 319

to residues at the interface of each L-subunit and two S-subunit βA-βB loops (Figure 5b). 320

Ala-228-Ser substitutions influence structural movements in these loops and can influence 321

kinetics via long range effects 38,39. Catalytic roles for Ser-328 and Glu-470 appear more 322

obvious. Amino acid 328 is located at the hinge of loop 6 that closes over the catalytic site 323

to facilitate intra-molecular interactions that influences both the fixation rate and partiality 324

for carboxylation or oxygenation40. Loop 6 closure involves the L-subunit C-terminus 325

where amino acid 470 resides (Fig 5b). As a hydrophobic Ala-470 in the Paniacea NAD-326

ME and C3 Rubisco, burial of the side chains into the enzyme surface may slow C-terminus 327

movement. In contrast, Glu/Gln-470 might enhance solvent exposure and increase C-328

terminal tail mobility to alter the dynamics of loop 6 closure and stimulate kcatc. 329

We sought to test the possible role of L-subunit amino acid replacement(s) in 330

influencing the variability in kinetics and temperature response among Paniceae Rubisco 331

by tobacco chloroplast transformation. Multiple chloroplast genome (plastome) 332

transformed tobacco lines were made (tobPdL and tobPbL) where the tobacco plastome rbcL 333

gene was replaced with the rbcL gene from P. bisulcatum or P. deustum were generated 334

(Fig S7a). Each transformed line was unable to survive outside of tissue culture (Fig 5c). 335

Despite producing ample levels of Panicum rbcL mRNA (Fig S7b), no hybrid L8S8 336

holoenzyme (comprising Panicum L-subunits and tobacco S-subunits, Fig S7c) or 337

unassembled Panicum L-subunits (Fig S7d) were detected. This suggests there are 338

incompatibilities in the biogenesis requirements (translation, folding and/or assembly) of 339

Rubisco between monocot and dicot species. 340

Discussion: 341

As calls for expanding the range of Rubiscos included in catalytic diversity studies 342

increase, so should the range of temperatures examined. Unlike prior C3-focused Rubisco 343

diversity studies, our high resolution catalytic screen revealed variation in the kinetic 344

trajectories of Paniceae Rubisco capable of enhancing C3-photosynthesis at temperatures 345

otherwise missed, or misjudged, from assaying at 25°C and one or two other temperatures. 346

Our analyses validate the co-evolution of higher kcatc and KC across C4-Rubiscos in 347

response to a CCM4,8,12,16,18,20, and unveil the widest variability in temperature kinetics 348

reported for vascular plant Rubisco to date. We uniquely reveal alignment of Rubisco 349

kinetics with CCM biochemistry and Paniceae biogeography. For example, the higher kcatc 350

and KC of the NADP-ME and PCK Paniceae Rubisco correlated with the forecast higher 351

BSC CO2 levels in these C4-subtypes relative to the NAD-ME and the CCM deficient C3/C2 352

species41. The slower decline in SC/O by PCK and NADP-ME Rubisco under increasing 353

temperature (Fig 3e) may reflect their warmer origins relative to the drier and cooler origins 354

of NAD-ME and C3 grasses, respectively42. Endeavors to determine whether these 355

correlations extend to other C4-species should take heed of inaccurately extrapolating the 356

response of kcatc to temperature using a single Arrhenius fit rather than correctly accounting 357

for its biphasic response that deviates at ~25°C (Fig 3b) – a relationship recognized 40 358

years ago32, but whose mechanistic origin remains an unsolved. 359

The clustering of carboxylase properties of Paniceae Rubisco according to 360

photosynthetic physiology contrasted with the more variable oxygenase activities (Fig 1i 361

& j) supporting assertions these competing reactions can evolve independently due to 362

differences in the mechanism and energy profile of their multi-step reactions23. This 363

variability engenders natural kinetic diversity which, on the Rubisco superfamily scale, is 364

relatively restricted for C3-Rubisco2,14,17,21,25,29 19,31. In contrast the broad kinetic diversity 365

among Paniceae Rubisco presents opportunities for enhancing C3-photosynthesis under 366

varying atmospheric CO2 and temperature (Fig S5). In particular P. bisulcatum (C3) and P. 367

deustum (PCK) Rubisco could distinctly improve C3-photosynthetic potential under cooler 368

and warmer temperatures, respectively, relative to the standardized tobacco Rubisco 369

control (Fig 4) #wheat/rice. The simulated improvements stemmed from temperature 370

dependent enhancements in SC/O and/or kcatc / KC

21%O2 (Fig. 3d,e), and not necessarily from 371

high kcatc (as emphasized by8). Our findings suggest that improving C3-crop photosynthesis 372

under warmer future climates may be best served by exploring the Rubisco kinetic diversity 373

of C4-land plants, in particular among PCK and NADP-ME species. 374

Four L-subunit residues could contribute kinetic diversity among Paniceae 375

Rubisco. These included two amino acids within structural regions whose movements 376

influence Rubisco kinetics: the catalytic loop 6 (residue 328) and C-terminal tail (residue 377

470). Positive selection of Ala-328-Ser substitutions have been reported for some 378

Limonium haplotypes43 and a few C3 and CAM plant species17. This suggests the higher 379

kcatc and KC of Paniceae NADP-ME and PCK Rubisco might arise from the Glu/Gln-470 380

substitution (Fig 5a). Our attempts to test this by heterologous expression in tobacco 381

chloroplasts proved unsuccessful (Fig 5c). The transformation limitation appears 382

associated with differences in the ancillary protein requirements of Paniceae Rubisco 383

biogenesis (Fig 5C), a constraint also preventing the production of Rubisco from red 384

algae44 and seemingly other monocot species45 in tobacco chloroplasts. 385

Our data indicate the S-subunits also likely influence the kinetic diversity among 386

Paniceae Rubisco. A comparable kinetic determining property was postulated for S-387

subunits in rice and wheat Rubisco34,35. In P. virgatum four RbcS mRNAs are made 388

(Phytozome) whose translated 121-123 amino acid S-subunits vary by 1 to 6 residues. 389

Mutagenic study of the multiple RbcS transcripts produced in Paniceae would be a 390

significant undertaking, but one possibly made easier using modern site specific nucleus 391

gene editing tools that are now available in a variety of crop species46. Clarifying the 392

influence of S-subunits on the temperature kinetics of Rubisco from differing plant origins 393

is critical to developing appropriate L- and/or S-subunit mutagenic technologies for 394

modifying crop Rubisco kinetics to suit future climates. 395

Corresponding author for material requests is spencer.whitney@anu.eud.au 396

Acknowledgements 397

We thank Asaph Cousins for supplying their S. viridis Rubisco kinetic data for analysis. 398

This research was funded by the following grants from the Australian Research Council: 399

DE130101760 (RES), DP120101603 (OG, SMW) and CE140100015 (OG, SMW). 400

Author contributions 401

RS, OG and SMW designed the study and undertook the experimental work. MVK 402

undertook the phylogenetic analysis and LHG the structural analysis. All authors 403

contributed to drafting the paper. 404

Competing financial interests 405

The authors declare no competing financial interests. 406

References 407

1 Ort, D. R. et al. Redesigning photosynthesis to sustainably meet global food and 408

bioenergy demand. Proc Natl Acad Sci U S A 112, 8529-8536 (2015). 409

2 Zhu, X. G., Long, S. P. & Ort, D. R. Improving photosynthetic efficiency for greater 410

yield. Annu Rev Plant Biol 61, 235-261 (2010). 411

3 Long, Stephen P., Marshall-Colon, A. & Zhu, X.-G. Meeting the global food 412

demand of the future by engineering crop photosynthesis and yield potential. Cell 413

161, 56-66 (2015). 414

4 Carmo-Silva, E., Scales, J. C., Madgwick, P. J. & Parry, M. A. J. Optimizing 415

Rubisco and its regulation for greater resource use efficiency. Plant Cell Env 38, 416

1817-1832 (2015). 417

5 Evans, J. R. Improving photosynthesis. Plant Physiol 162, 1780-1793 (2013). 418

6 Parry, M. A. J. et al. Rubisco activity and regulation as targets for crop 419

improvement. J Exp Bot 64, 717-730 (2013). 420

7 von Caemmerer, S., Quick, W. P. & Furbank, R. T. The development of C4 rice: 421

current progress and future challenges. Science 336, 1671-1672 (2012). 422

8 Whitney, S. M., Houtz, R. L. & Alonso, H. Advancing our understanding and 423

capacity to engineer nature's CO2-sequestering enzyme, Rubisco. Plant Physiol 424

155, 27-35 (2011). 425

9 Andersson, I. Catalysis and regulation in Rubisco. J Exp Bot 59, 1555-1568 (2008). 426

10 Andersson, I. & Backlund, A. Structure and function of Rubisco. Plant Physiol 427

Biochem 46, 275-291 (2008). 428

11 Sharwood, R. E., Ghannoum, O. & Whitney, S. M. Prospects for improving CO2 429

fixation in C3-crops through understanding C4-Rubisco biogenesis and catalytic 430

diversity. Currt Opin Plant Biol 31, 135-142 (2016). 431

12 Carmo-Silva, A. E. et al. Rubisco activities, properties, and regulation in three 432

different C4 grasses under drought. J Exp Bot 61, 2355-2366 (2010). 433

13 Galmes, J. et al. Expanding knowledge of the Rubisco kinetics variability in plant 434

species: environmental and evolutionary trends. Plant Cell Environ 37, 1989-2001 435

(2014). 436

14 Galmés, J., Kapralov, M. V., Copolovici, L. O., Hermida-Carrera, C. & Niinemets, 437

Ü. Temperature responses of the Rubisco maximum carboxylase activity across 438

domains of life: phylogenetic signals, trade-offs, and importance for carbon gain. 439

Photosynth Res 123, 183-201 (2015). 440

15 Jordan, D. B. & Ogren, W. L. The CO2/O2 specificity of ribulose 1,5-bisphosphate 441

carboxylase oxygenase - dependence on ribulosebisphosphate concentration, pH 442

and temperature. 161, 308-313 (1984). 443

16 Prins, A. et al. Rubisco catalytic properties of wild and domesticated relatives 444

provide scope for improving wheat photosynthesis. J Exp Bot 67, 1827-1838 445

(2016). 446

17 Young, J. N. et al. Large variation in the Rubisco kinetics of diatoms reveals 447

diversity among their carbon-concentrating mechanisms. J Exp Bot 67, 3445-3456 448

(2016). 449

18 Andrews, T. J. & Whitney, S. M. Manipulating ribulose bisphosphate 450

carboxylase/oxygenase in the chloroplasts of higher plants. Arch.Biochem.Biophys. 451

414, 159-169 (2003). 452

19 Raven, J. A. Rubisco: still the most abundant protein of Earth? New Phytol 198, 1-453

3 (2013). 454

20 Sage, R. F. The evolution of C4 photosynthesis New Phytol 161, 341-370 (2004). 455

21 Sage, R. F., Christin, P.-A. & Edwards, E. J. The C4 plant lineages of planet Earth. 456

J Exp Bot 62, 3155-3169 (2011). 457

22 Furbank, R. T. Evolution of the C4 photosynthetic mechanism: are there really three 458

C4 acid decarboxylation types? J Exp Bot 62, 3103-3108 (2011). 459

23 Sage, R. F., Sage, T. L. & Kocacinar, F. Photorespiration and the evolution of C4 460

photosynthesis. Ann Rev Plant Biol 63, 19-47 (2012). 461

24 Ghannoum, O. et al. Faster rubisco is the key to superior nitrogen-use efficiency in 462

NADP-malic enzyme relative to NAD-malic enzyme C4 grasses. Plant Physiol 137, 463

638-650 (2005). 464

25 Sharwood, R., von Caemmerer, S., Maliga, P. & Whitney, S. The catalytic 465

properties of hybrid Rubisco comprising tobacco small and sunflower large 466

subunits mirror the kinetically equivalent source Rubiscos and can support tobacco 467

growth. Plant Physiol 146, 83-96 (2008). 468

26 Sharwood, R. E., Sonawane, B. V., Ghannoum, O. & Whitney, S. M. Improved 469

analysis of C4 and C3 photosynthesis via refined in vitro assays of their carbon 470

fixation biochemistry. J Exp Bot 67, 3137-3148 (2016). 471

27 Whitney, S. M. & Sharwood, R. E. Construction of a tobacco master line to improve 472

Rubisco engineering in chloroplasts. J Exp Bot 59, 1909-1921 (2008). 473

28 Tcherkez, G. G. B., Farquhar, G. D. & Andrews, T. J. Despite slow catalysis and 474

confused substrate specificity, all ribulose bisphosphate carboxylases may be 475

nearly perfectly optimized. Proc Nat Acad Sci 103, 7246-7251 (2006). 476

29 Boyd, R. A., Gandin, A. & Cousins, A. B. Temperature response of C4 477

photosynthesis: biochemical analysis of Rubisco, phosphoenolpyruvate 478

carboxylase and carbonic anhydrase in Setaria viridis. Plant Physiol 169, 1850-479

1861 (2015). 480

30 Perdomo, J. A., Cavanagh, A. P., Kubien, D. S. & Galmés, J. Temperature 481

dependence of in vitro Rubisco kinetics in species of Flaveria with different 482

photosynthetic mechanisms. Photosynth Res 124, 67-75 (2015). 483

31 Sage, R. F. Variation in the kcat of Rubisco in C3 and C4 plants and some 484

implications for photosynthetic performance at high and low temperature. J Exp 485

Bot 53, 609-620 (2002). 486

32 Savir, Y., Noor, E., Milo, R. & Tlusty, T. Cross-species analysis traces adaptation 487

of Rubisco toward optimality in a low-dimensional landscape. Proc Nat Acad Sci 488

107, 3475-3480 (2010). 489

33 Pearcy, R. W. & Ehleringer, J. Comparative ecophysiology of C3 and C4 plants. 490

Plant Cell Env 7, 1-13 (1984). 491

34 Tcherkez, G. The mechanism of Rubisco-catalyzed oxygenation. Plant Cell Env 492

39, 983-997 (2016). 493

35 Farquhar, G. D., von Caemmerer, S. & Berry, J. A. A biochemical model of 494

photosynthetic CO2 assimilation in leaves of C3 species. Planta 149, 78-90 (1980). 495

36 Sharwood, R. E. & Whitney, S. M. Correlating Rubisco catalytic and sequence 496

diversity within C3 plants with changes in atmospheric CO2 concentrations. Plant 497

Cell Env 37, 1981-1984 (2014). 498

37 Sharkey, T. D., Bernacchi, C. J., Farquhar, G. D. & Singsaas, E. L. Fitting 499

photosynthetic carbon dioxide response curves for C3 leaves. Plant Cell Env 30, 500

1035-1040 (2007). 501

38 Walker, B., Ariza, L. S., Kaines, S., Badger, M. R. & Cousins, A. B. Temperature 502

response of in vivo Rubisco kinetics and mesophyll conductance in Arabidopsis 503

thaliana: comparisons to Nicotiana tabacum. Plant Cell Env 36, 2108-2119 (2013). 504

39 Hermida-Carrera, C., Kapralov, M. V. & Galmés, J. Rubisco catalytic properties 505

and temperature response in crops. Plant Physiol., doi:10.1104/pp.16.01846 506

(2016). 507

40 Orr, D. et al. Surveying Rubisco diversity and temperature response to improve 508

crop photosynthetic efficiency. Plant Physiol., doi:10.1104/pp.16.00750 (2016). 509

41 Badger, M. R. & Collatz, G. J. Studies on the kinetic mechanism of RuBP 510

carboxylase and oxygenase reactions, with particular reference to the effect of 511

temperature on kinetic papameters. Carnegie YB 76, 355-361 (1977). 512

42 Ishikawa, C., Hatanaka, T., Misoo, S., Miyake, C. & Fukayama, H. Functional 513

incorporation of sorghum small subunit increases the catalytic turnover rate of 514

Rubisco in transgenic rice Plant Physiol 156, 1603-1611 (2011). 515

43 Hauser, T., Popilka, L., Hartl, F. U. & Hayer-Hartl, M. Role of auxiliary proteins 516

in Rubisco biogenesis and function. Nat Plants 1 (2015). 517

44 Wachter, R. M. et al. Activation of interspecies-hybrid Rubisco enzymes to assess 518

different models for the Rubisco-Rubisco activase interaction. Photosynth Res 117, 519

557-566 (2013). 520

45 Spreitzer, R. J., Peddi, S. R. & Satagopan, S. Phylogenetic engineering at an 521

interface between large and small subunits imparts land-plant kinetic properties to 522

algal Rubisco. Proc Natl Acad Sci 102, 17225-17230 (2005). 523

46 von Caemmerer, S. & Furbank, R. T. The C4 pathway: an efficient CO2 pump. 524

Photosynth Res 77, 191-207, doi:10.1023/a:1025830019591 (2003). 525

47 Still, C. J., Pau, S. & Edwards, E. J. Land surface skin temperature captures thermal 526

environments of C3 and C4 grasses. Glob Ecol Biogeo 23, 286-296 (2014). 527

48 Galmés, J. et al. Environmentally driven evolution of Rubisco and improved 528

photosynthesis and growth within the C3 genus Limonium (Plumbaginaceae). New 529

Phytol 203, 989-999 (2014). 530

49 Whitney, S. M. & Andrews, T. J. Plastome-encoded bacterial ribulose-1, 5-531

bisphosphate carboxylase/oxygenase (RubisCO) supports photosynthesis and 532

growth in tobacco. Proc Nat Acad Sci 98, 14738-14743 (2001). 533

50 Bortesi, L. & Fischer, R. The CRISPR/Cas9 system for plant genome editing and 534

beyond. Biotech Adv 33, 41-52 (2015). 535

536

Figure Legends 537

538

Figure 1. The diversity in the catalytic properties of Rubisco at 25°C across C3 and 539

C4 grasses within Paniceae. 540

Box plots of comparative (a) leaf dry matter 12C/13C isotopic fractionation (δ13C) and (b to 541

f) in vitro measured Rubisco kinetics from tobacco and Paniceae species with C2, C3 and 542

varying C4 subtypes (NADP ME, NAD ME and PCK). See Table S1 for species list. 543

Median values shown in boxes as vertical line, 95% confidence limits represented by 544

horizontal lines. Letter variation indicates significant differences (p < 0.05) between 545

parameters (Table S1). Kinetic properties analyzed include (b) substrate saturated 546

carboxylation and (d) oxygenation turnover rates (kcatc, kcat

o), the Michaelis constants (Km) 547

for (c) CO2 and (e) O2 (KC, KO) and (f) relative specificity for CO2 over O2 (SC/O). (g to j) 548

Pairwise relationships among the kinetic parameters to assess the quality of their linear 549

correlations (dashed line). 550

551

Figure 2. Variation among the Paniceae Rubisco kinetics differentially affect 552

simulated rates of C3-photosynthesis at 25°C. 553

Comparison of the relationships between kcatc and either (a) SC/O or (b) KC

21%O2, the value 554

for Kc under ambient O2 (O) calculated as KC(1+O/KO) (Table S1). The r2 values show the 555

quality of their linear correlations (dashed lines). (c) The influence of the averaged 556

Paniceae C3 and C4 subtype Rubisco kinetics (Table S1) on CO2 assimilation rates (A) at 557

25°C in a C3-leaf as a function of Cc. Lines are modelled (Farquhar et al., (1980)) using the 558

carboxylase activity limited assimilation equation: 559

𝐴 =(𝐶𝑐. 𝑠𝑐 − 0.5 𝑂/𝑆𝑐/𝑜) 𝑘𝑐𝑎𝑡

𝑐 . 𝐵

𝐶𝑐. 𝑠𝑐 + 𝐾(1 + 𝑂 𝐾𝑜⁄ )− 𝑅𝑑 560

using a CO2 solubility in H2O (sc) of 0.0334M bar-1, an O of 253 µM, a Rubisco content 561

(B) of 30 µmol catalytic sites.m2 and a non-photorespiratory CO2 assimilation rate (Rd) of 562

1 µmol.m-2.s-1. The light limited CO2 assimilation rates (to the right of the symbols) were 563

modelled according to the equation: 564

𝐴 =(𝐶𝑐. 𝑠𝑐 − 0.5 𝑂/𝑆𝑐/𝑜) 𝐽

4(𝐶𝑐 . 𝑠𝑐 + 𝑂/𝑆𝑐/𝑜)− 𝑅𝑑 565

assuming an electron transport rate (J) of 160 µmol.m-2.s-1. Yellow shading indicates where 566

the modeled CO2-assimilation rates of C3, NADP-ME and PCK Paniceae Rubisco exceed 567

that of Rubisco from the model C3-plant, tobacco (dotted line). 568

569

570

Figure 3. Divergence in the catalytic properties of Paniceae and tobacco Rubisco in 571

response to temperature. 572

(a) The heat of activation (ΔHa) and scaling constant (c) for the kinetic parameters of 573

tobacco Rubisco and the mean (±S.E) values measured for the various Paniceae species 574

with C4 (NAD ME, NADP ME, PCK) or C3 (including the aligning C2) biochemical 575

physiologies (see Table S3). Letters show the statistical ranking using a post hoc Tukey 576

test among the biochemical physiology groupings (different letters indicated differences at 577

the 5% level, p < 0.05). (b to e) Differences in the temperature response of tobacco (grey 578

dashed line) and the averaged kinetic properties for Rubisco from Paniceae species with 579

varying biochemical physiologies. The lines are derived as described in Figures S1 to S4 580

using the values listed in panel (a). 581

582

583

Figure 4. The potential for improving the thermal response of C3-photosynthesis. 584

The benefits of (a) P. bisulcatum (C3) and (b) P. deustum (C4-PCK) Rubisco to the rate of 585

photosynthesis in a C3-leaf under varying chloroplast CO2 concentrations (Ca) and 586

temperature (see scale). Rate increases are presented as a percentage above that provided 587

by tobacco Rubisco. The data was modelled according to24 using the parameters listed in 588

Table S4 and plotted in Fig S4. 589

590

591

Figure 5. Approaches to decipher possible catalytic switches in the L-subunit of 592

Paniceae Rubisco. 593

(a) Amino acid variation in the L-subunit of each Paniacea Rubisco analysed in this study. 594

A phylogenetic analysis of the L-subunit sequences and their Genbank accession 595

information is provided in Figure S6. (b) Structure of spinach L8S8 Rubisco (L-subunits in 596

green, S-subunits grey) viewed from the top (left) and side (middle) showing the relative 597

locations of 94Glu on the solvent-exposed Rubisco surface (yellow triangle), 228Ala in the 598

α2 helix (orange triangle), 328Ser at the hinge of loop 6 (purple triangle) and 470Pro in the 599

C-terminal tail extension (blue triangle) of one L-subunit. A closer view of a L-subunit pair 600

(right) with one showing ribbon structural detail and the other showing the positioning of 601

94Glu, 228Ala, 328Ser and 470Pro relative to each other, an N-terminal domain loop, the α2 602

helix, loop 6, C-terminal tail extension and S-subunit βA-βB loops (yellow, orange, purple, 603

blue and grey, respectively). An active site bound reaction-intermediate analogue 2-CABP 604

is shown as a ball and stick. (c) Chloroplast transformation of the Rubisco L-subunit genes 605

from P. bisulcatum (Pbis-rbcL) and P. deustum (Pbis-rbcL) into tobacco was undertaken 606

to identify the amino acids (catalytic switches) responsible for their differing catalytic 607

properties. No Rubisco biogenesis was detected in the tobPbL and tobPdL tobacco genotypes 608

produced. Accordingly these plants could only grow in tissue culture on sucrose containing 609

media and were highly chlorotic (as shown). Detailed analysis of the transformation, 610

Rubisco mRNA and protein biochemistry is provided in Figure S7. 611

612

Supplemental data 613

Variation in response of C3 and C4 Paniceae Rubisco to temperature 614

provides opportunities for improving C3 photosynthesis 615

Robert E. Sharwood1+, Oula Ghannoum2+*, Maxim V. Kapralov1,3, Laura H. Gunn1,4, and 616

Spencer M. Whitney1+*. 617

1Research School of Biology, Australian National University, Canberra ACT, 2601, 618

Australia. 619

2 Hawkesbury Institute for the Environment, Western Sydney University, Richmond 620

NSW, 2753, Australia. 621

+ ARC Centre of Excellence for Translational Photosynthesis, Australian National 622

University Canberra ACT, 2601, Australia. 623

3Current Address: School of Natural Sciences and Psychology, Liverpool John Moores 624

University, Liverpool, L3 3AF, United Kingdom. 625

4Current address: Department of Cell and Molecular Biology, Uppsala University, 626

Uppsala, SE-751 24, Sweden. 627

*Corresponding Authors: o.ghannoum@westernsydney.edu.au and 628

spencer.whitney@anu.eud.au 629

630

Number of Supplemental Figures: 7 631

Number of Supplemental Tables: 4 632

633

Figure S1. Variation in the temperature response of kcatc among the Paniceae and tobacco 634

Rubisco. 635

(a) Substrate saturated rates of carboxylation (kcatc) determined using soluble leaf protein extract (n ≥ 3 636

biological samples per species) were measured at 10, 15, 20, 25, 30 and 37°C with the expected 637

exponential decrease in kcatc less evident at the lower assay temperatures for all Rubisco samples, as 638

evident in prior published data 18-20,28,30-32 }, including that of Boyd et al., (2015) 15 (orange triangles) for 639

Seteria viridis Rubisco measured by Membrane Inlet Mass Spectrometry (MIMS). Plotted data are listed 640

in Table S2. (b) Evaluation of the data via Arrhenius-style plots (i.e. ln kcatc vs 1/T) indicated the kcat

c 641

response diverged at around 25°C. Shown are the averaged linear fits to each Arrhenius plot in a biphasic 642

manner with the < 25°C measurements (i.e. at 10, 15, 20, 25°C) separated from the > 25°C measurements 643

(25, 30 and 37°C). The averaged data values were fitted to the following equation 644

𝑃𝑎𝑟𝑎𝑚𝑒𝑡𝑒𝑟 = 𝑒𝑥𝑝 [𝑐 −∆𝐻𝑎

𝑅𝑇] 645

and the heat of activation (ΔHa) for kcatc at both < 25°C and > 25°C was derived from the slope (ln(kcat

c) 646

= -ΔHa/R; where R is the molar gas constant, 8.314 J K-1 mol-1) and the scaling constant (c) from the 647

ordinal intercept. The calculated values are listed in Table S3 and were fitted to the above equation to 648

derive the exponential curves in panel a. For comparison, the fitted lines for tobacco Rubisco kcatc data 649

are shown as dotted lines in each C4 Rubisco plot. See Table S3 for statistical analysis. 650

651

652

Figure S2. Variation in the temperature response of CO2 affinity among the Paniceae and tobacco 653

Rubisco. 654

(a) The Michealis constant for CO2 measured in the presence of ambient (253 µM) O2 concentration 655

(KC21%O2) determined from the same assays used to determine kcat

c in Fig S1 (n ≥ 3 biological samples 656

analyzed per specie) varied exponentially over 10 to 37°C. Orange triangles, data for Seteria viridis 657

Rubisco measured by MIMS15). Plotted data are listed in Table S2. (b) Arrhenius-style plots of the data 658

with the averaged linear regression fitted as described in Fig S1 to determine the heat of activation (ΔHa) 659

and the scaling constant (c) values listed in Table S3 and used to derive the exponential curves shown in 660

panel (a). As a scaling comparison, the fitted lines for tobacco Rubisco KC21%O2 data are shown as dotted 661

lines in each C4 Rubisco plot. See Table S3 for statistical analysis. 662

663

664

Figure S3. Variation in the temperature response of CO2 over O2 specificity among the Paniceae 665

and tobacco Rubisco 666

(a) The influence of temperature on the specificity of CO2 over O2 (SC/O) determined using Rubisco 667

purified from at least two biological samples per specie. Orange triangles, data for Seteria viridis Rubisco 668

measured by MIMS15. Plotted data are listed in Table S2. (b) Arrhenius plots of the data with the 669

averaged linear regression fitted as described in Fig S1 to determine the heat of activation (ΔHa) and the 670

scaling constant (c) values for SC/O listed in Table S3 and used to derive the exponential curves shown in 671

panel (a). The fitted lines for tobacco Rubisco SC/O data are shown as dotted lines in each C4 Rubisco 672

plot as a scaling comparison. See Table S3 for statistical analysis. 673

674

Figure S4. Variation in carboxylation efficiency under ambient O2 among the Paniceae and 675

tobacco Rubisco under varying temperature. 676

The temperature response of carboxylation efficiency (kcatc/ KC

21%O2) for each Rubisco was calculated by 677

dividing the averaged values of kcatc (FigS1) by its corresponding KC

21%O2 values (Fig S2) for each assay 678

temperature. As shown previously for Flaveria 19 and Seteria viridis 15 Rubisco, the carboxylation 679

efficiency of each Paniceae and tobacco Rubisco declined at temperatures below 25°C and showed less 680

variation at temperatures > 25°C. Shown are the linear regressions fitted to the averaged Paniceae data 681

for each biochemical physiology. See Table S3 for statistical analysis. 682

683

Figure S5. Effect of Rubisco kinetics on the thermal photosynthetic response. 684

The effects Rubisco catalytic properties on the thermal response of leaf photosynthesis (A) to leaf chloroplastic CO2 concentration (Cc). 685

The curves were modelled according to Farquhar et al. (1980) 24 using equations and parameters shown in Supplementary Table S5. The 686

solid and dashed lines refer to the Rubisco limited (Ac) and RuBP-regeneration limited (Aj) assimilation rates, respectively. The circles 687

refer to assimilation rates under current Ca (400 bar, white) and that predicted for 2050 (550 bar, black). Data for tobacco Rubisco 688

shown in grey in each panel for comparison. 689

690

Figure S6. Rubisco L-subunit phylogeny in the Paniceae. 691

Maximum likelihood phylogeny of Rubisco L-subunit sequences from the fourteen Paniceae species 692

examined in this study relative to the outgrouped Rubisco from Hordeum vulgare (barley) and Triticum 693

aestivum (wheat). ML trees assembled under the Dayhoff model implemented in RAxML v.8 47 using 694

translated L-subunit sequences from the full length rbcL genes available from the following Genbank 695

accession: P. bisulcatum, (*); S. laxa, (*); P. milioides, (*); P. antidotale, (*); P. monticola, (*); C. 696

ciliaris, (*); S. viridis, (KT289405.1); P. virgatum, (HQ731441.1); P. milliaceum , (KU343177.1); P. 697

coloratum, (*); M. maximus, (*); U. panicoides, (*); U. panicoides , (*); P. deustum , (*); U. 698

mosambicensis, (*); H. vulgare, (KT962228.1) and T. aestivum, (KJ592713.1). *sequences submitted to 699

Genbank, awaiting accession numbers. 700

701

Figure S7. Chloroplast transformation of the P. bisulcatum (C3) and P. deustum (C4-PCK) rbcL 702

genes to assess Paniceae Rubisco biogenesis in tobacco. 703

(a) Comparison of the plastome sequence in wild-type, cmtrL and the plastome transformed tobPbL and 704

tobPdL tobacco genotypes generated in this study. Duplicate tobPbL and tobPdL lines were made by 705

plastome transformation as described 13 by homologous recombination replacement of the cmrbcM gene 706

in the plastome of the cmtrL tobacco genotype with rbcL genes for P. bisulcatum or P. deustum Rubisco 707

(synthesized to match the tobacco rbcL nucleotide sequence where feasible) and the aadA selectable 708

marker gene (coding resistance to spectinomycin). Numbering represents the flanking plastome 709

sequence in the pLEVPdL and pLEVPbL transforming plasmids. P, 292-bp rbcL promoter/5’UTR; T, 710

288-bprbcL 3’UTR; T, 112-bp of psbA 3’UTR; t, 147-bp rps16 3’UTR. Position of the 221-bp 5UTR 711

probe 13 and the corresponding rbcL and rbcL-aadA mRNAs (dashed lines) to which it hybridizes are 712

indicated. (b) Total leaf RNA (5µg) extracted from tissue culture grown plant samples was separated on 713

denaturing formaldehyde gels and the EtBr stained RNA visualised (upper panel) before blotting onto 714

Hybond-N nitrocellulose membrane (GE healthcare) as described 11 and probed with the 32P-labelled 715

5UTR probe (lower panel). The probe correctly hybridised to the wild-type tobacco rbcL mRNA and the 716

rbcL and rbcL-aadA mRNA transcripts in each tobPbL and tobPdL line. (c) Soluble leaf protein from the 717

same leaves analyzed in (b) was processed for measuring Rubisco levels by NdPAGE analysis and 14C-718

CABP quantification as described 25. While wildtype tobacco L8S8 Rubisco was readily detected by 14C-719

CABP binding, Coomassie staining and by immunoblot analysis with an antibody to tobacco Rubisco 720

following ndPAGE, these methods detected no hybrid L8S8 Rubisco biogenesis in the tobPbL and tobPdL 721

genotypes (i.e. complexes comprising the introduce Panicum L-subunits and the endogenous, cytosol 722

made tobacco S-subunits). (d) Further inspection of the soluble and total (comprising soluble + insoluble) 723

leaf protein separated by SDS PAGE did not detect any Rubisco L-subunit (~50 kDa) or S-subunit (~14.5 724

kDa) in either cellular protein fraction of the tobPbL or tobPdL lines by Comassie staining or Rubisco 725

antibody blot analysis. This indicated that even when grown in tissue culture the resource limitations 726

confronting the photosynthetically deplete tobPbL and tobPdL lines precluded the synthesis and /or 727

accumulation of Panicum sp. Rubisco L-subunits. Whether co-expressing their cognate SSu or/and Raf1 728

48 can circumvent this biogenesis challenge remains to be tested. 729

Table S1: Summary of the catalytic parameters of Paniceae and tobacco Rubisco at 25°C. 730

Species Physiology

δ13C kcatc KC KC

21% O2 kcato KO kcat

o/KO SC/O kcat

c/

KC21%O2

kcatc/

KC

(‰) (s-1) (µM) (µM) (s-1) (µM) (mM-1.s-1) (mol.mol-1) (mM.-1.s-1) (mM.-1.s-1)

Pa

nic

eae

spec

ies

Panicum antidotale

C4 NADP-ME

-12..92 3.9 ± 0.2 n.m 25.2 n.m n.m n.m 74.5 ± 0.4 156 n.m

Panicum monticola -13.53 5.3 ± 0.7 18.2 ± 0.5 26.6 2.0 543 ± 67 3.7 79.4 ± 1.7 198 290

Cenchrus ciliaris -12.46 6.0 ± 0.8 19.0 ± 0.7 29.2 2.1 470 ± 52 4.5 69.9 ± 3.0 205 314

Setaria viridis -13.81 5.9 ± 0.5 18.1 ± 0.6 25.5 2.8 619 ± 86 4.4 72.7 ± 0.2 230 323

Panicum virgatum C4

NAD-ME

-13.98 3.3 ± 0.9 12.7 ± 0.1 24.5 0.9 271 ± 12 3.1 82.6 ± 2.8 133 258

Panicum milliaceum -15.50 2.1 ± 0.3 7.2 ± 0.3 13.1 1.1 313 ± 46 3.6 79.9 ± 4.3 159 287

Panicum coloratum -14.20 3.4 ± 0.6 11.1 ± 0.5 17.3 1.6 445 ± 58 3.6 84.8 ± 2.8 197 308

Megathyrsus maximus

C4 PCK

-14.32 5.3 ± 0.5 13.9 ± 0.8 27.1 1.3 265 ± 34 4.7 80.3 ± 2.8 195 380

Urochloa panicoides -14.51 5.6 ± 0.6 15.4 ± 0.7 24.1 2.1 444 ± 80 4.6 78.3 ± 0.3 232 364

Panicum deustum -12.62 5.0 ± 0.5 15.4 ± 0.2 28.1 1.2 306 ± 16 3.8 84.8 ± 0.2 177 322

Urochloa mosambicensis -13.08 5.7 ± 0.7 14.8 ± 0.4 22.8 2.2 464 ± 79 4.7 82.5 ± 1.3 252 388

Panicum milioides C2 -31.50 2.2 ± 0.3 7.4 ± 0.3 12.1 1.3 387 ± 46 3.3 92.3 ± 1.0 182 301

Panicum bisulcatum

C3

-28.68 2.6 ± 0.4 7.8 ± 0.3 12.6 1.6 416 ± 67 3.8 87.7 ± 1.5 207 333

Steinchisma laxa -28.70 2.3 ± 0.3 7.7 ± 0.5 12.4 1.4 419 ± 89 3.2 91.4 ± 4.8 184 294

Nicotiana tabacum n.m. 3.1 ± 0.3 9.7 ± 0.1 18.3 1.1 283 ± 15 3.9 82.6 ± 0.8 168 318

Averages of

parameters for

Paniceae Rubisco

C3 -28.7±0.1 a 2.4±0.2 a 7.8±0.1 a 12.5±0.1 a 1.5±0.1 a 418±1 a 3.5±0.3 a 90±2 a 195±12 a 313±19 ab

NAD-ME -14.6±0.5 b 2.9±0.4 a 10.3±1.6 a 18.3±3.3 a 1.2±0.2 a 343±52 a 3.4±0.2 a 82±1 a 163±18 a 284±15 a

NADP-ME -13.3±0.4 b 5.7±0.2 b 18.4±0.3 c 27.1±1.1 b 2.3±0.2 a 544±43 a 4.2±0.3 a 74±3 b 211±10 a 309±10 ab

PCK -13.6±0.5 b 5.4±0.2 b 14.9±0.4 b 25.5±1.2 b 1.7±0.3 a 370±49 a 4.5±0.2 a 81±1 a 214±17 a 363±15 b

Type (p) *** *** *** ** ns(0.08) ns(0.072) * ** ns *

For each species data are the mean±SE of at least N=3 biological samples assayed in duplicate. One-way ANOVA was undertaken 731

using the photosynthetic type/subtype as the main factor. Symbols show the statistical significance levels (ns = p > 0.05; * = p < 0.05; 732

** = p < 0.01; ***: p < 0.001), while letters show the ranking of the means using a post hoc Tukey test (different letters indicate 733

statistical differences at the 5% level, p < 0.05). kcato, maximal oxygenation rate calculated from SC/O = (kcat

c /KC)/(kcato /KO). KC

21%O2, 734

KC under ambient atmospheric O2 levels (O = 252 µM O2 in air saturated H2O) calculated as KC(1+O/KO). n.m, not measured. 735

Table S2: The catalytic parameters of Paniceae and tobacco Rubisco between 10°C and 37°C. 736 Temp tobacco P. bis. P. milioi.

C2/C3

P. miliac P. color P. virg NAD

ME

P. mont. S. virid C. cilaris NADP

ME

P. deust M. max U. panic

PCK (°C)

N = 4

(x2)

N = 3

(x2)

N = 3

(x2)

N = 4

(x2)

N = 3

(x2)

N = 3

(x2)

N = 4

(x2)

N = 3

(x2)

*Boyd et.

al 2015

N = 3

(x2)

N = 4

(x2)

N = 3

(x2)

N = 3

(x2)

kcatC (s-1)

10 0.88±0.19 0.53±0.04 0.49±0.04 0.51±0.02 0.44±0.01 0.97±0.06 0.96±0.08 0.79±0.17 1.52±0.06 1.70±0.03 0.88±0.08 1.81±0.14 1.68±0.08 1.40±0.02 1.75±0.04 1.58±0.11 1.58±0.10

15 1.37±0.22 1.09 ±0.07 0.92±0.15 1.00 ±0.08 0.79±0.06 1.70±0.09 1.48±0.24 1.32±0.27 2.64±0.27 2.80±0.15 1.55±0.23 2.99±0.17 2.81±0.10 2.21±0.05 2.80±0.10 2.75±0.14 2.59±0.19

20 1.98±0.13 1.47±0.02 1.34±0.07 1.40±0.07 1.26 ±0.01 2.43±0.24 2.27±0.29 1.99±0.37 3.54±0.37 3.88±0.24 3.47±0.17 4.15±0.31 3.86±0.18 3.42±0.31 3.65±0.02 3.64±0.50 3.57±0.08

25 3.10±0.08 2.60±0.19 2.21±0.09 2.41±0.20 2.08±0.05 3.41±0.09 3.30±0.06 2.93±0.43 5.28±0.16 5.85±0.33 5.21±0.22 5.98±0.09 5.70±0.21 4.96±0.32 5.38±0.12 5.60±0.26 5.28±0.18

30 3.78±0.37 2.83±0.07 2.87±0.41 2.85±0.02 2.37±0.12 3.70±0.11 3.87±0.26 3.31±0.48 6.29±0.51 6.65±0.29 8.33±2.82 7.04±0.87 6.66±0.22 5.80±0.28 6.34±0.13 6.59±0.51 6.25±0.23

37 5.45±0.67 4.26±0.30 4.15±0.47 4.20±0.06 3.24±0.24 4.86±0.77 5.34±0.90 4.48±0.64 8.49±1.03 8.78±0.28 13.88±1.35 9.01±1.21 8.76±0.15 7.64±0.43 7.76±0.16 8.51±0.71 7.97±0.27

(35°C)

KC21%O2 (µM)

10 7.6±1.9 5.9±0.1 7.7±2.2 6.8±0.9 4.9±1.0 10.4±1.5 11.8±3.9 9.0±2.1 11.0±2.1 19.8±2.3 35.4±5.5 18.6±0.4 16.5±2.8 14.1±2.8 17.8±0.5 11.7±1.1 14.6±1.8

15 11.0±1.4 9.3±2.6 8.4±0.4 8.8±0.5 9.8±0.6 15.2±1.3 19.7±1.4 14.9±2.8 15.8±4.5 24.6±3.2 37.9±6.7 26.1±0.8 22.2±3.2 19.5±2.1 21.6±0.3 18.2±0.4 19.8±1.0

20 12.9±2.0 8.3±0.6 9.2±1.2 8.8±0.5 7.5±0.2 13.2±1.6 15.9±1.0 12.2±2.5 18.4±4.3 20.0±2.4 50.6±7.0 23.6±1.5 20.7±1.5 20.1±2.8 18.8±0.3 16.6±0.1 18.5±1.0

25 18.3±0.9 12.6±1.0 12.1±0.8 12.4±0.3 13.1±1.1 17.3±1.7 24.5±2.1 18.3±3.3 26.6±4.0 25.5±2.0 57.5±6.3 29.2±2.3 27.1±1.1 28.1±2.0 27.1±1.1 24.1±1.0 26.4±1.2

30 22.0±3.4 13.6±1.2 14.4±1.4 14.0±0.4 15.7±2.5 19.1±2.3 30.1±3.7 21.9±4.6 31.7±4.8 30.6±2.0 103.8±9.4 35.4±5.4 32.6±1.5 32.5±2.3 33.9±2.3 28.2±2.2 31.5±1.7

37 30.8±3.8 20.6±5.3 19.2±01.5 19.9±0.8 19.6±5.3 25.1±6.0 45.9±3.3 30.2±8.0 45.8±5.4 39.3±3.6 138.1±15.3 45.6±7.2 43.5±2.1 42.0±3.0 39.2±2.0 38.0±6.6 39.7±1.2

(35°C)

kcatC / KC

21%O2 (mM-1.s-1)

10 111 90 64 77±13 90 93 81 88±4 138 86 26 97 107±16 99 98 135 111±12

15 125 117 110 114±3 81 111 75 89±11 167 114 42 115 132±17 113 129 152 131±11

20 154 178 144 161±17 167 185 142 165±12 192 194 73 176 189±6 170 194 219 194±14

25 169 206 183 194±12 159 197 135 164±18 198 229 93 205 211±9 177 195 232 201±16

30 172 208 200 203±4 151 194 125 156±20 198 217 81 199 205±6 179 187 234 200±17

37 177 206 217 211±5 165 194 116 158±23 186 223 104 198 202±11 182 198 224 201±12

(35°C)

SC/O (mol.mol-1)

N = 4

(x3)

N = 2

(x3)

N = 2

(x3)

N = 2

(x3)

N = 3

(x3)

N = 2

(x3)

N = 2

(x3)

N = 2

(x3)

N = 3

(x3)

N = 2

(x3)

N = 2

(x3)

10 130.0±1.0 152.7±2.0 152.4±4.9 152.5±0.1 127.1±2.6 138.2±5.1 142.7±2.6 136.0±4.7 n.m. 115.9±3.3 74.9±3.6 106.4±0.8 111.2±3.9 128.0±6.4 119.0±2.9 128.6±2.0 125.2±3.1

15 114.7±2.1 126.1±2.4 120.8±0.8 123.4±2.7 114.7±2.6 109.3±0.8 117.5±1.4 113.8±2.4 108.0±2.2 74.5±9.1 93.2±0.8 100.6±6.0 117.7±2.0 103.4±3.3 106.5±1.6 109.2±4.3

20 94.9±0.4 102.0±0.6 105.1±1.2 103.5±1.6 105.7±8.4 90.1±1.9 95.4±0.5 97.0±4.6 79.1±0.5 64.1±7.9 79.9±0.1 79.5±0.3 100.3±1.3 90.7±0.5 86.7±2.4 92.6±4.0

25 82.0±0.6 87.7±1.5 92.3±1.0 90.0±2.3 80.0±4.3 78.8±0.8 82.6±3.3 80.4±1.1 72.7±0.2 61.5±2.5 69.9±2.0 71.3±1.2 84.8±0.2 80.3±2.8 78.2±0.3 81.1±1.9

30 69.0±2.1 68.6±1.4 67.7±1.4 68.2±0.4 67.0±0.6 70.2±0.9 74.3±0.2 70.5±2.1 64.2±1.1 35.6±2.0 60.6±0.1 62.4±1.3 75.7±0.3 60.9±2.4 69.0±0.6 68.5±4.3

35 61.1±0.7 53.7±3.2 58.1±1.7 55.9±2.2 46.1±3.4 59.4±0.3 61.2±1.1 55.6±4.8 47.1±2.2 27.8±3.4 52.8±0.5 49.9±2.3 62.9±0.7 56.9±0.8 58.0±1.4 59.3±1.8

Rubisco catalysis parameters (average ± S.E.) measured in duplicate (x2) or triplicate (x3) for each biological sample (N). *Data for S. 737

viridis Rubisco shown in blue italic (N=4 ± S.E.) measured by MIMS 15. The data for each species are plotted in Figures S1 to S4 with 738

the parameter averages (± SE) for tobacco Rubisco and for each Paniceae photosynthetic type/subtype shown in bold and shaded grey 739

are plotted in Figures 3. n.m, not measured. The data from this study are statistically analyzed following derivation of the heat of 740

activation (ΔHa) and scaling constant (c) values for each parameter (see Table S3). 741

Table S3. Heat of activation (ΔHa) and scaling constant (c) values among Paniceae and tobacco Rubisco kinetic parameters. 742

743

Species

Physiology

kcatc (>25°C) kcat

c (<25°C) KC21% O2 SC/O

ΔHa (± S.D) c (± S.D) ΔHa (± S.D) c (± S.D) ΔHa (± S.D) c (± S.D) ΔHa (± S.D) c (± S.D)

kJ.mol-1 kJ.mol-1 kJ.mol-1 kJ.mol-1

Nicotiana tabacum C3

36.4±2.9 15.8±1.2 60.3±1.9 25.5±0.8 37.3±1.4 17.9±0.7 22.5±1.2 -4.7±0.3

Panicum bisulcatum 32.4±8.5 13.9±3.7 71.2±7.2 29.7±3.0 31.0±3.7 15.0±1.8 29.7±2.0 -7.6±0.5

Panicum milioides C2 40.3±0.5 17.0±0.2 68.4±4.2 28.4±1.8 25.4±1.9 12.7±1.0 27.6±3.0 -6.7±0.7

Panicum monticola

C4 (NADP-ME

30.6±1.8 14.0±0.8 56.5±4.2 24.5±1.8 33.8±1.9 16.9±0.9 n.m

Cenchrus ciliaris 26.3±0.7 12.4±0.3 54.9±3.1 23.9±1.3 22.1±2.4 12.4±1.3 20.3±0.5 -4.0±0.1

Setaria viridis 26.3±2.8 12.4±1.3 56.6±2.8 24.6±1.2 21.4±0.1 11.9±0.1 25.3±4.4 -6.0±1.0

#Boyd et al 2015 75.0±1.1 31.9±0.5 94.2±10.4 39.3±9.4 44.7±3.8 22.3±1.9 25.2±5.7 -6.2±1.4

Panicum virgatum

C4 (NAD-ME)

31.1±2.9 13.7±1.3 58.1±0.8 24.7±0.3 34.0±4.5 16.9±2.2 23.8±1.9 -5.2±0.4

Panicum milliaceum 28.6±4.0 12.3±1.7 71.6±1.8 29.6±0.7 34.9±5.8 16.6±2.7 28.6±6.1 -7.2±1.5

Panicum coloratum 23.2±4.5 10.6±2.1 58.3±4.6 24.8±1.9 21.3±2.8 11.4±1.5 23.5±2.3 -5.1±0.5

Megathyrsus maximus

C4 (PCK)

24.5±1.2 11.6±0.5 50.2±3.3 21.9±1.4 22.2±3.0 12.3±1.7 22.2±3.3 -4.6±0.7

Urochloa panicoides 26.9±0.9 12.6±0.4 57.3±4.5 24.8±1.9 29.9±3.0 15.2±1.5 22.3±2.0 -4.6±0.4

Panicum deustum 27.8±1.7 12.8±0.8 59.5±1.3 25.6±0.5 28.7±1.7 14.9±0.9 20.8±2.3 -3.9±0.4

Averages C3/C2 36.3 b 15.5 b 69.8 b 29.0 b 28.2 a 13.9 a 28.6 b -7.1 a

NAD-ME 27.6 ab 12.2 a 62.7 ab 26.6 ab 30.0 a 15.0 a 25.3 ab -5.8 ab

NADP-ME 27.7 a 12.9 a 56.0 a 24.4 a 25.8 a 13.7 a 22.8 a -5.0 b

PCK 26.4 a 12.3 a 55.6 a 24.1 a 26.9 a 14.1 a 21.7 a -4.4 b

Type (p) ns(0.081) ns(0.10) * ns(0.062) ns ns ns(0.072) ns(0.089)

Values of ΔHa and c were determined from measures of kcatc (Fig S1), KC

21%O2 (Fig S2) and SC/O (Fig S4) made at 10, 15, 20, 25, 30 and 35 (or 744

37)°C (see Table S2 for data) and fitted to the Arrhenius-type equation 745

𝑃𝑎𝑟𝑎𝑚𝑒𝑡𝑒𝑟 = 𝑒𝑥𝑝 [𝑐 −∆𝐻𝑎

𝑅𝑇] 746

where R the molar gas constant (8.314 J K−1 mol−1) and T the assay temperature (K). n.m, not measured. For each species data are the mean ± SE 747

of at least N=3 biological samples assayed in duplicate. One-way ANOVA was undertaken using the photosynthetic type/subtype as the main 748

factor. Symbols show the statistical significance levels (ns = p > 0.05; * = p < 0.05), while letters show the ranking of the means using a post hoc 749

Tukey test (different letters indicate statistical differences at the 5% level, p < 0.05). # Comparative data for S. viridis Rubisco from Boyd et 750

al., (2015) 15 measured by MIMS. 751

Table S4. Summary of parameters used in the modelling plots shown in Figure 4 and Figure S5. 752

Temperature (oC)

Species Parameters 10 15 20 25 30 37 Reference

Nicotiana

tabacum

kcatc,(s-1) 1.4 1.8 2.4 3.1 4.7 8.0 Tables S1-2

KCair (µM) 7.9 10.4 13.6 17.6 22.5 31.4 Tables S1-2

SC/O (M M-1) 131 111 95 81 70 57 Tables S1-2

Jmax/Vcmax 2.6 2.1 1.8 1.6 1.4 1.2 Sharkey et al. (2007) 27

Panicum

bisulcatum

kcatc,(s-1) 1.2 1.6 2.0 2.6 4.2 8.0 Tables S1-2

KCair (µM) 6.2 7.8 9.7 12.1 14.8 19.6 Tables S1-2

SC/O (M M-1) 156 125 101 83 68 52 Tables S1-2

Panicum

deustum

kcatc,(s-1) 2.7 3.3 4.1 5.1 7.6 12.9 Tables S1-2

KCair (µM) 14.6 18.1 22.2 27.1 32.8 42.4 Tables S1-2

SC/O (M M-1) 132 114 98 85 74 61 Tables S1-2

Urochloa

panicoides

kcatc,(s-1) 3.2 3.8 4.7 5.6 8.3 13.9 Tables S1-2

KCair (µM) 12.5 15.5 19.2 23.6 28.8 37.6 Tables S1-2

SC/O (M M-1) 125 106 91 78 67 55 Tables S1-2

Common

parameters

gm (mol m-2 s-1 bar-1) 0.24 0.33 0.44 0.57 0.72 0.97 von Caemmerer and Evans (2015) 49

TPU (µmol m-2 s-1) 3.81 5.62 8.14 11.40 14.65 14.10 Sharkey et al. (2007) 27

Rd (µmol m-2 s-1) 0.37 0.52 0.73 1.04 1.36 2.06 Sharkey et al. (2007) 27

sc (M bar-1) 0.0512 0.0442 0.0383 0.0334 0.0292 0.0245 https://en.wikipedia.org/wiki/Henry's_law

so (M bar-1) 0.00170 0.00154 0.00139 0.00126 0.00115 0.00101 https://en.wikipedia.org/wiki/Henry's_law

Jmax (µmol m-2 s-1) 63 87 118 160 214 317 Sharkey et al. (2007) 27

Rubisco sites (µmol s-1) 30 30 30 30 30 30

Photosynthesis rate, A was calculated as A = min (Ac, Aj, At), where Ac, Aj and At are the CO2-limited (Ac), light-limited (Aj) and the triose 753

phosphate utilisation (TPU)-limited (At) assimilation rates, respectively. Their expressions are defined as: 754

𝐴𝑐 =𝑚.𝑘cat

𝑐(𝐶𝑐.𝑠𝑐−0.5𝑂𝑐 𝑆𝑐 𝑜⁄⁄ )

(𝐶𝑐.𝑠𝑐+𝐾𝑐air)

− 𝑅𝑑 ; 755

𝐴𝑗 =(𝐶𝑐.𝑠𝑐−0.5𝑂𝑐 𝑆𝑐 𝑜⁄⁄ )𝐽max

4(𝐶𝑐.𝑠𝑐+𝑂𝑐 𝑆𝑐 𝑜⁄⁄ )− 𝑅𝑑 ; and 756

𝐴𝑡 = 3TPU − 𝑅𝑑 . 757

The model parameters used are: m = amount of leaf Rubisco set at 30 mol active sites m-2; kcatc (s-1) = Rubisco carboxylation rate; KC

air 758

(M) = Michaelis-Menten constant of Rubisco for CO2 and SC/O = CO2/O2 specificity of Rubisco. The maximal RuBP carboxylation-759

limited assimilation rate, Vcmax = m.kcatc. The maximal RuBP regeneration-limited assimilation rate, Jmax (mol m-2 s-1) is set to equal 760

1.7Vcmax for tobacco at 25oC; its values at other temperatures were calculated using the thermal dependence from Bernacchi et al (2003) 761

50: 762

𝐽𝑚𝑎𝑥(𝑇) = 𝐽𝑚𝑎𝑥25∙ 𝑒

(𝑐−𝐻𝑎

𝑅.(273+𝑇)), where c = 17.7 and Ha = 43.9 (in kJ mol-1). 763

The values at 25oC for TPU (11.4 mol m-2 s-1) and mitochondrial respiration, Rd (1 mol m-2 s-1) and their thermal dependence were 764

adapted from Sharkey et al (2007) 27: 765

𝑅𝑑(𝑇) = 𝑅𝑑25∙ 𝑒

(𝑐−𝐻𝑎

𝑅.(273+𝑇)), where c = 18.72 and Ha = 46.4 (in kJ mol-1) 766

𝑇𝑃𝑈(𝑇) = 𝑇𝑃𝑈25 [𝑒

(𝑐−𝐻𝑎

𝑅.(273+𝑇))

1+𝑒(

𝑆.(273+𝑇).𝐻𝑑𝑅.(273+𝑇)

)], where c =21.46 and Ha = 53.1, Hd = 201.8 and 0.65 (in kJ mol-1). 767

Cc and Oc are the CO2 and O2 concentrations in the chloroplast, respectively. Gas concentrations in the liquid phase were calculated 768

using the solubility constants for CO2 (sc = 0.0334 M bar-1) and O2 (so = 0.00126 M bar-1) at 25oC. Their thermal dependence was 769

determined according to Henry’s law using the following expressions (https://en.wikipedia.org/wiki/Henrys_law): 770

𝑠𝑐(𝑇) = 𝑠𝑐25∙ 2400. 𝑒(

1

273+𝑇−

1

298) and 771

𝑠𝑜(𝑇) = 𝑠𝑜25∙ 1700. 𝑒(

1273+𝑇

−1

298) 772

Intercellular CO2 concentration, Ci was calculated using a constant Ci/Ca ratio of 0.7049. Cc was calculated as Cc = Ci - A/gm, where gm 773

is the mesophyll conductance to CO2 transfer. Tobacco gm at 25oC (0.57 mol m-2 s-1 bar-1) and its thermal dependence were taken from 774

von Caemmerer and Evans (2015) 49. 775

54

Table S5. Parameters used to calculate CO2 concentrations in 14CO2-fixation 776

assays at the varying temperatures. 777

Parameter (units) Value

T; assay temperature (10°C) (15°C) (20°C) (25°C) (30°C) (37°C)

283K 288K 293K 298K 303K 310K

q; CO2 solubility at 1 atm (Mol.L-1.atm-1) 0.0524 0.0455 0.0382 0.0329 0.0289 0.0240

R: universal gas constant (L.atm.K-1.mol-1) 0.082057

pK1 6.362 6.327 6.280 6.251 6.226 6.202

pK2 10.499 10.431 10.377 10.329 10.290 10.238

*pH 8.27 8.24 8.21 8.16 8.11 8.03

The values were fitted to the Henderson-Hasselbalch derived equation 778

[𝐶𝑂2] =(𝐶𝑡)

1 +𝑉

𝑣𝑞𝑅𝑇 + 10(𝑝𝐻−𝑝𝐾1) + 10(2𝑝𝐻−𝑝𝐾1−𝑝𝐾2) 779

V/v: ratio of reaction vial headspace (V) to assay volume (v). 780

*example pH variation for 50 mM EPPES-NaOH buffer adjusted to pH 8.16 at 25°C; the 781

0.26 pH variation has <1% effect on tobacco Rubisco carboxylase activity. 782

55

References 783

1 Andrews, T. J. & Whitney, S. M. Manipulating ribulose bisphosphate 784

carboxylase/oxygenase in the chloroplasts of higher plants. Arch.Biochem.Biophys. 785

414, 159-169 (2003). 786

2 Carmo-Silva, E., Scales, J. C., Madgwick, P. J. & Parry, M. A. J. Optimizing 787

Rubisco and its regulation for greater resource use efficiency. Plant Cell Env 38, 788

1817-1832 (2015). 789

3 Raven, J. A. Rubisco: still the most abundant protein of Earth? New Phytol 198, 1-790

3 (2013). 791

4 Sage, R. F. The evolution of C4 photosynthesis New Phytol 161, 341-370 (2004). 792

5 Sage, R. F., Christin, P.-A. & Edwards, E. J. The C4 plant lineages of planet Earth. 793

J Exp Bot 62, 3155-3169 (2011). 794

6 Furbank, R. T. Evolution of the C4 photosynthetic mechanism: are there really three 795

C4 acid decarboxylation types? J Exp Bot 62, 3103-3108 (2011). 796

7 Sage, R. F., Sage, T. L. & Kocacinar, F. Photorespiration and the evolution of C4 797

photosynthesis. Ann Rev Plant Biol 63, 19-47 (2012). 798

8 Sharwood, R. E., Ghannoum, O. & Whitney, S. M. Prospects for improving CO2 799

fixation in C3-crops through understanding C4-Rubisco biogenesis and catalytic 800

diversity. Currt Opin Plant Biol 31, 135-142 (2016). 801

9 Ghannoum, O. et al. Faster rubisco is the key to superior nitrogen-use efficiency in 802

NADP-malic enzyme relative to NAD-malic enzyme C4 grasses. Plant Physiol 137, 803

638-650 (2005). 804

10 von Caemmerer, S., Quick, W. P. & Furbank, R. T. The development of C4 rice: 805

current progress and future challenges. Science 336, 1671-1672 (2012). 806

11 Sharwood, R., von Caemmerer, S., Maliga, P. & Whitney, S. The catalytic 807

properties of hybrid Rubisco comprising tobacco small and sunflower large 808

subunits mirror the kinetically equivalent source Rubiscos and can support tobacco 809

growth. Plant Physiol 146, 83-96 (2008). 810

12 Sharwood, R. E., Sonawane, B. V., Ghannoum, O. & Whitney, S. M. Improved 811

analysis of C4 and C3 photosynthesis via refined in vitro assays of their carbon 812

fixation biochemistry. J Exp Bot 67, 3137-3148 (2016). 813

13 Whitney, S. M. & Sharwood, R. E. Construction of a tobacco master line to improve 814

Rubisco engineering in chloroplasts. J Exp Bot 59, 1909-1921 (2008). 815

14 Tcherkez, G. G. B., Farquhar, G. D. & Andrews, T. J. Despite slow catalysis and 816

confused substrate specificity, all ribulose bisphosphate carboxylases may be 817

nearly perfectly optimized. Proc Nat Acad Sci 103, 7246-7251 (2006). 818

15 Boyd, R. A., Gandin, A. & Cousins, A. B. Temperature response of C4 819

photosynthesis: biochemical analysis of Rubisco, phosphoenolpyruvate 820

carboxylase and carbonic anhydrase in Setaria viridis. Plant Physiol 169, 1850-821

1861 (2015). 822

16 Carmo-Silva, A. E. et al. Rubisco activities, properties, and regulation in three 823

different C4 grasses under drought. J Exp Bot 61, 2355-2366 (2010). 824

17 Galmes, J. et al. Expanding knowledge of the Rubisco kinetics variability in plant 825

species: environmental and evolutionary trends. Plant Cell Environ 37, 1989-2001 826

(2014). 827

56

18 Galmés, J., Kapralov, M. V., Copolovici, L. O., Hermida-Carrera, C. & Niinemets, 828

Ü. Temperature responses of the Rubisco maximum carboxylase activity across 829

domains of life: phylogenetic signals, trade-offs, and importance for carbon gain. 830

Photosynth Res 123, 183-201 (2015). 831

19 Perdomo, J. A., Cavanagh, A. P., Kubien, D. S. & Galmés, J. Temperature 832

dependence of in vitro Rubisco kinetics in species of Flaveria with different 833

photosynthetic mechanisms. Photosynth Res 124, 67-75 (2015). 834

20 Sage, R. F. Variation in the kcat of Rubisco in C3 and C4 plants and some 835

implications for photosynthetic performance at high and low temperature. J Exp 836

Bot 53, 609-620 (2002). 837

21 Savir, Y., Noor, E., Milo, R. & Tlusty, T. Cross-species analysis traces adaptation 838

of Rubisco toward optimality in a low-dimensional landscape. Proc Nat Acad Sci 839

107, 3475-3480 (2010). 840

22 Pearcy, R. W. & Ehleringer, J. Comparative ecophysiology of C3 and C4 plants. 841

Plant Cell Env 7, 1-13 (1984). 842

23 Tcherkez, G. The mechanism of Rubisco-catalyzed oxygenation. Plant Cell Env 843

39, 983-997 (2016). 844

24 Farquhar, G. D., von Caemmerer, S. & Berry, J. A. A biochemical model of 845

photosynthetic CO2 assimilation in leaves of C3 species. Planta 149, 78-90 (1980). 846

25 Sharwood, R. E. & Whitney, S. M. Correlating Rubisco catalytic and sequence 847

diversity within C3 plants with changes in atmospheric CO2 concentrations. Plant 848

Cell Env 37, 1981-1984 (2014). 849

26 Young, J. N. et al. Large variation in the Rubisco kinetics of diatoms reveals 850

diversity among their carbon-concentrating mechanisms. J Exp Bot 67, 3445-3456 851

(2016). 852

27 Sharkey, T. D., Bernacchi, C. J., Farquhar, G. D. & Singsaas, E. L. Fitting 853

photosynthetic carbon dioxide response curves for C3 leaves. Plant Cell Env 30, 854

1035-1040 (2007). 855

28 Walker, B., Ariza, L. S., Kaines, S., Badger, M. R. & Cousins, A. B. Temperature 856

response of in vivo Rubisco kinetics and mesophyll conductance in Arabidopsis 857

thaliana: comparisons to Nicotiana tabacum. Plant Cell Env 36, 2108-2119 (2013). 858

29 Whitney, S. M., Houtz, R. L. & Alonso, H. Advancing our understanding and 859

capacity to engineer nature's CO2-sequestering enzyme, Rubisco. Plant Physiol 860

155, 27-35 (2011). 861

30 Hermida-Carrera, C., Kapralov, M. V. & Galmés, J. Rubisco catalytic properties 862

and temperature response in crops. Plant Physiol., doi:10.1104/pp.16.01846 863

(2016). 864

31 Orr, D. et al. Surveying Rubisco diversity and temperature response to improve 865

crop photosynthetic efficiency. Plant Physiol., doi:10.1104/pp.16.00750 (2016). 866

32 Badger, M. R. & Collatz, G. J. Studies on the kinetic mechanism of RuBP 867

carboxylase and oxygenase reactions, with particular reference to the effect of 868

temperature on kinetic papameters. Carnegie YB 76, 355-361 (1977). 869

33 Jordan, D. B. & Ogren, W. L. The CO2/O2 specificity of ribulose 1,5-bisphosphate 870

carboxylase oxygenase - dependence on ribulosebisphosphate concentration, pH 871

and temperature. 161, 308-313 (1984). 872

57

34 Ishikawa, C., Hatanaka, T., Misoo, S., Miyake, C. & Fukayama, H. Functional 873

incorporation of sorghum small subunit increases the catalytic turnover rate of 874

Rubisco in transgenic rice Plant Physiol 156, 1603-1611 (2011). 875

35 Prins, A. et al. Rubisco catalytic properties of wild and domesticated relatives 876

provide scope for improving wheat photosynthesis. J Exp Bot 67, 1827-1838 877

(2016). 878

36 Hauser, T., Popilka, L., Hartl, F. U. & Hayer-Hartl, M. Role of auxiliary proteins 879

in Rubisco biogenesis and function. Nat Plants 1 (2015). 880

37 Wachter, R. M. et al. Activation of interspecies-hybrid Rubisco enzymes to assess 881

different models for the Rubisco-Rubisco activase interaction. Photosynth Res 117, 882

557-566 (2013). 883

38 Andersson, I. & Backlund, A. Structure and function of Rubisco. Plant Physiol 884

Biochem 46, 275-291 (2008). 885

39 Spreitzer, R. J., Peddi, S. R. & Satagopan, S. Phylogenetic engineering at an 886

interface between large and small subunits imparts land-plant kinetic properties to 887

algal Rubisco. Proc Natl Acad Sci 102, 17225-17230 (2005). 888

40 Andersson, I. Catalysis and regulation in Rubisco. J Exp Bot 59, 1555-1568 (2008). 889

41 von Caemmerer, S. & Furbank, R. T. The C4 pathway: an efficient CO2 pump. 890

Photosynth Res 77, 191-207, doi:10.1023/a:1025830019591 (2003). 891

42 Still, C. J., Pau, S. & Edwards, E. J. Land surface skin temperature captures thermal 892

environments of C3 and C4 grasses. Glob Ecol Biogeo 23, 286-296 (2014). 893

43 Galmés, J. et al. Environmentally driven evolution of Rubisco and improved 894

photosynthesis and growth within the C3 genus Limonium (Plumbaginaceae). New 895

Phytol 203, 989-999 (2014). 896

44 Whitney, S. M. & Andrews, T. J. Plastome-encoded bacterial ribulose-1, 5-897

bisphosphate carboxylase/oxygenase (RubisCO) supports photosynthesis and 898

growth in tobacco. Proc Nat Acad Sci 98, 14738-14743 (2001). 899

45 Parry, M. A. J. et al. Rubisco activity and regulation as targets for crop 900

improvement. J Exp Bot 64, 717-730 (2013). 901

46 Bortesi, L. & Fischer, R. The CRISPR/Cas9 system for plant genome editing and 902

beyond. Biotech Adv 33, 41-52 (2015). 903

47 Stamatakis, A. RAxML version 8: a tool for phylogenetic analysis and post-904

analysis of large phylogenies. Bioinform 30, 1312-1313 (2014). 905

48 Whitney, S. M., Birch, R., Kelso, C., Beck, J. L. & Kapralov, M. V. Improving 906

recombinant Rubisco biogenesis, plant photosynthesis and growth by coexpressing 907

its ancillary RAF1 chaperone. Proc Natl Acad Sci 112, 3564-3569 (2015). 908

49 von Caemmerer, S. & Evans, J. R. Temperature responses of mesophyll 909

conductance differ greatly between species. Plant Cell Env 38, 629-637 (2015). 910

50 Bernacchi, C. J., Pimentel, C. & Long, S. P. In vivo temperature response functions 911

of parameters required to model RuBP-limited photosynthesis. Plant Cell Env 26, 912

1419-1430 (2003). 913

914

915